Date of Award

Summer 8-1-2015

Degree Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Dr. Yongsheng Xu

Second Advisor

Dr. Vivi Alatas

Third Advisor

Dr. Carlianne Patrick

Fourth Advisor

Dr. Sally Wallace


This dissertation explores the information asymmetry problem between government and applicants of means-tested welfare benefits. The first chapter examines the problem in a static setting. To meet benefit eligibility, individuals reduce their labor supply so that they meet the required income level. This action enables them to increase their payoff but makes it difficult for government to differentiate the needy from the non-needy. Using a signaling game model and incorporating income-related ordeals (i.e. required-hour and essential component) as the signals, this study finds that government can mitigate the problem, increase social welfare, and maintain income-based eligibility requirements.

The second chapter investigates the effect of the asymmetry when benefit applicants have choices of where and when to apply, and the solution to the asymmetry problem. Jurisdiction differential in net benefit may induce a migration from a low to a high net-benefit jurisdiction. Likewise, individuals may choose to procrastinate on the application process if they find that applying later gives higher net-benefit than applying now. Consequently, these behaviors create information asymmetry. Using a signaling game model, the result shows that, after accounting for required-hour and essential component, the individuals have no incentive to migrate or procrastinate on the application process in order to be eligible.

The third chapter evaluates the solution to the problem when maintaining a particular take-up rate is also a government’s objective. Imposing ordeal in the welfare system is expected to help government direct the benefit to the needy but this may reduce the take-up rate. Conditioning on any particular take-up rate that the government aims to achieve, the models provide a solution on how to distinguish the types of applicants without harming the take-up rate. Using 2013 U.S. state level data, the model predicts that adjusting the cutoff level of the ordeals may change the take-up rates anywhere between 0.008 and 9.233 percent. This range represents the total number of marginal individuals who are in the programs. This sheds some light on what particular cutoff level of ordeals a government should impose so that it does not harm the targeted take-up rates.